Hi Mahnaz, Why don't you follow the advice of the edgeR User's Guide (as Mark has suggested)? All the case studies in the User's Guide describe how the filtering was done in a principled way. Total count filtering is not so bad, but it is susceptible to being driven by one library, especially by one library with a large sequence depth. The procedure described by Mark and used in the guide is a compromise of several considerations. BTW, there are newer versions of R and edgeR available than what you are using. Best wishes Gordon > Date: Wed, 30 Apr 2014 21:34:50 +0200 > From: Mark Robinson <mark.robinson at="" imls.uzh.ch=""> > To: "Ryan C. Thompson" <rct at="" thompsonclan.org=""> > Cc: bioconductor at r-project.org, Mahnaz Kiani <mahnazkiani at="" gmail.com=""> > Subject: Re: [BioC] total count filter cutoff > > > In my lab, we typically follow a "CPM of at least X in at least Y > samples" rule, where X=1 (arbitrary but reasonable, can be changed) and > Y=size of smallest replicate group, according to one of the case studies > in the user's guide, for example: > > ------ > 4.3.6 Filtering > We filter out very lowly expressed tags, keeping genes that are > expressed at a reasonable level in at least one treatment condition. > Since the smallest group size is three, we keep genes that achieve at > least one count per million (cpm) in at least three samples: > >> keep <- rowSums(cpm(y)>1) >= 3 >> y <- y[keep,] > ------ > > (http://www.bioconductor.org/packages/release/bioc/vignettes/edgeR/i nst/doc/edgeRUsersGuide.pdf) > > Cheers, Mark > > > ---------- > Prof. Dr. Mark Robinson > Statistical Bioinformatics, Institute of Molecular Life Sciences > University of Zurich > http://ow.ly/riRea > Date: Wed, 30 Apr 2014 11:29:28 -0700 (PDT) > From: "mahnaz Kiani [guest]" <guest at="" bioconductor.org=""> > To: bioconductor at r-project.org, mahnazkiani at gmail.com > Subject: [BioC] total count filter cutoff > > > I'm using edgeR for analysis of may data and I'm not sure what total > count filter value cutoff value I should use, My reads are paired 50bP > reads and total reads per sample is about 80,000,000. I tried cutoff > values of 5,10,15,30,50 and 100 and I only saw differences between 50 > and 100 but still looking for logical reason to chose the cutoff value. > > Appreciate your help, > Mahnaz > > -- output of sessionInfo(): > > R 3.0.2 ______________________________________________________________________ The information in this email is confidential and intend...{{dropped:4}}

Mahnaz, Just one more comment: with the large sequence depth that you have, you can afford to go down to a low cpm cutoff in order to include very lowly expressed genes and transcripts in your analysis. You could try cpm>0.2 or cpm>0.1. Best Gordon On Fri, 2 May 2014, Gordon K Smyth wrote: > Hi Mahnaz, > > Why don't you follow the advice of the edgeR User's Guide (as Mark has > suggested)? All the case studies in the User's Guide describe how the > filtering was done in a principled way. > > Total count filtering is not so bad, but it is susceptible to being driven by > one library, especially by one library with a large sequence depth. The > procedure described by Mark and used in the guide is a compromise of several > considerations. > > BTW, there are newer versions of R and edgeR available than what you are > using. > > Best wishes > Gordon > > >> Date: Wed, 30 Apr 2014 21:34:50 +0200 >> From: Mark Robinson <mark.robinson at="" imls.uzh.ch=""> >> To: "Ryan C. Thompson" <rct at="" thompsonclan.org=""> >> Cc: bioconductor at r-project.org, Mahnaz Kiani <mahnazkiani at="" gmail.com=""> >> Subject: Re: [BioC] total count filter cutoff >> >> >> In my lab, we typically follow a "CPM of at least X in at least Y samples" >> rule, where X=1 (arbitrary but reasonable, can be changed) and Y=size of >> smallest replicate group, according to one of the case studies in the >> user's guide, for example: >> >> ------ >> 4.3.6 Filtering > >> We filter out very lowly expressed tags, keeping genes that are expressed >> at a reasonable level in at least one treatment condition. Since the >> smallest group size is three, we keep genes that achieve at least one count >> per million (cpm) in at least three samples: >> >>> keep <- rowSums(cpm(y)>1) >= 3 >>> y <- y[keep,] >> ------ >> >> (http://www.bioconductor.org/packages/release/bioc/vignettes/edgeR/ inst/doc/edgeRUsersGuide.pdf) >> >> Cheers, Mark >> >> >> ---------- >> Prof. Dr. Mark Robinson >> Statistical Bioinformatics, Institute of Molecular Life Sciences >> University of Zurich >> http://ow.ly/riRea > > >> Date: Wed, 30 Apr 2014 11:29:28 -0700 (PDT) >> From: "mahnaz Kiani [guest]" <guest at="" bioconductor.org=""> >> To: bioconductor at r-project.org, mahnazkiani at gmail.com >> Subject: [BioC] total count filter cutoff >> >> >> I'm using edgeR for analysis of may data and I'm not sure what total count >> filter value cutoff value I should use, My reads are paired 50bP reads and >> total reads per sample is about 80,000,000. I tried cutoff values of >> 5,10,15,30,50 and 100 and I only saw differences between 50 and 100 but >> still looking for logical reason to chose the cutoff value. >> >> Appreciate your help, >> Mahnaz >> >> -- output of sessionInfo(): >> >> R 3.0.2 > ______________________________________________________________________ The information in this email is confidential and intend...{{dropped:4}}